随着扩散模型（Diffusion models）的提出与迅速发展，依托其高度可解释的数学特性及高质量和多样性的结果，逐渐打破对抗生成网络(Generative Adversarial Network, GANs)在图像生成和图像编辑领域的垄断地位，基于扩散模型的图像编辑逐渐成为计算机视觉领域的研究热点。本文首先介绍了图像编辑的任务定义和扩散模型的基本原理；然后重点分类依次介绍了基于扩散模型的图像编辑技术的发展历程；总结了图像编辑领域常用的评价指标和数据集，同时定性定量比较了经典方法在不同数据集上的效果；最后对基于扩散模型的图像编辑现状进行总结和展望。

With the introduction and rapid development of diffusion models, these frameworks have begun to challenge the dominance of Generative Adversarial Networks (GANs) in the realms of image generation and editing, thanks to their highly interpretable mathematical properties and the high quality and diversity of their outputs. Image editing based on diffusion models is emerging as a research hotspot in the field of computer vision. This paper first introduces the task definition of image editing and the basic principles of diffusion models. It then categorizes and details the developmental trajectory of image editing techniques based on diffusion models. Furthermore, the paper reviews common evaluation metrics and datasets used in the image editing domain, and provides both qualitative and quantitative comparisons of classical methods across various datasets. Finally, it summarizes the current state and prospects of image editing based on diffusion models. Keywords: image editing; Computer Vision; diffusion model